sentimenTrader Blog


2017-09-21 | Eric Brown | Comments

Currently, there are 9 Stock indicators at Pessimistic Extremes and 38 Stock indicators at Optimistic Extremes. The tables below display stock market indicators that are at Very Optimistic, Extremely Optimistic, Very Pessimistic or Extremely Pessimistic levels. Note: These tables contain only indicators from the Stocks section of the site. For the tables below, you can click through to each chart […]

This post is available to sentimenTrader members only.

Please login to to view this content.

Don't have an account?

Sign up to get RISK-FREE access to all of our indicators, models, commentary and award-winning research.

If you've never tried the service before, then there is no charge for the first 30 days. Then pay as little as $1.59 per trading day for access to our award-winning research. There are three convenient billing options: $39/month, $109/quarter or $399/year.


2017-09-16 | Jason Goepfert | Comments

We’ve had some requests to post updates on a few popular models from the site, so below are current snapshots of the spread between our Smart Money and Dumb Money Confidence, the Medium-Term Optimism Index, and a proxy of the CNN Fear & Greed model.

All three reached modest levels of pessimism in mid-to-late August for the first time in a year. That proved to be enough for the dip-buyers as we’ve quickly rebounding right back to new highs. Because of the speed of the move, the models are showing that sentiment hasn’t had time to catch up to the price move, except for the CNN model which is just now entering “excess optimism” territory.

From our Backtest Engine, we can see that over the past few years, when the CNN model was in “excess optimism” territory, the S&P 500 averaged a return of 0.23% over the next month, below average for that time span. And it’s well below the 0.95% average return when the model was not in extreme territory (below 70).


2017-09-11 | Jason Goepfert | Comments

Over the past few months, financials have gotten hit hard. During the past week, fewer than 20% of members in the sector were trading above their 50-day moving averages, a deeply oversold reading, especially during the past 8 years of this bull market. That weakness has pushed Financials to be the weakest among major sectors. […]

This post is available to sentimenTrader members only.

Please login to to view this content.

Don't have an account?

Sign up to get RISK-FREE access to all of our indicators, models, commentary and award-winning research.

If you've never tried the service before, then there is no charge for the first 30 days. Then pay as little as $1.59 per trading day for access to our award-winning research. There are three convenient billing options: $39/month, $109/quarter or $399/year.


2017-08-28 | Eric Brown | Comments

Over the weekend we pushed out a new version of our backtesting engine (see more info about the engine here and here) that includes a chart for the index and indicator that you are backtesting as well as results delivered to the same page (so you won’t need to go back and forth between the backtest screen and results screen).

Now, when you go to the backtesting engine, you’ll see the following screen:

sentimenTrader backtesting engine with chart

Once you select an index  and an indicator, the chart will be drawn and will look similar to other charts on the site.

Once you’ve selected the index/indicator, the rest of the backtesting engine is the same. You can select a moving average (which will also draw on the chart) lookback period, observation period, indicator condition, indicator level and the option to exclude overlapping observations.

When ready to run the backtest, click the ‘Run Backtest’ button and the results will be rendered directly below the chart and backtest criteria section.  Additionally, if you want to run another backtest, just scroll back to the top of the page and change the parameters and click ‘Run backtest.’

sentimenTrader backtesting engine with completed chart

sentimenTrader backtesting engine with results


2017-08-25 | Jason Goepfert | Comments

With the introduction of our backtesting engine, it’s now possible to determine how a market (or almost any market) has responded to various conditions from our indicators.

Let’s go over a use case with the small-cap Russell 2000 ETF, IWM.

From the Dashboard, if we click on the Optix Heatmap and then reorganize it by the 20-day average of the Optimism Index, we can see that IWM is among the most hated of all markets over the past month.

If you click that IWM tile, it will bring up the chart of the Optimism Index for IWM. From the drop-down box below the chart, choose the 20-day average and it will show up on the chart. To isolate it, in the legend click IWM Optix, which will make everything disappear. Again in the legend, click 20-Period Moving Average and only the moving average will now appear in the chart. We can see that it’s just curling up from a low level of around 26.

Zoom out on the chart by using the window below the chart, or select 5Y from the boxes to have the chart re-scale and show the past 5 years of history. We can see that the current 20-day average is among the lowest we’ve seen during that time frame. To see how IWM has performed after similar extremes, click the Backtest This Indicator button.

The fund and indicator will already be chosen for you. Then select the 20-period moving average, for the past 8 years (regular subscribers will see history up to the past 5 years, premium tier members can test the whole history), let’s look at the next 40 days (not important), and how the market has done after the indicator “crossed above” the level of 27. Then click Run Backtest.

When the results pop up, click the Multi-Timeframe Results tab.

From there, we can see that during this bull market, when IWM has been as hated as it was recently over the past month and sentiment has begun to recover, a week later IWM actually sold off every time but once, and a few of them were stiff. But by two months later, it was higher every time, suggesting that buying into a short-term pullback has been a winning bet after sentiment started to recover from a pessimistic level.

This won’t work as well (or at all) during a bear market, but if you go back further you can see how it performed in 2007-2008 as well, to better determine what kind of risk was normal.


2017-08-05 | Eric Brown | Comments

We’ve just added a new ‘multi-timeframe’ feature to our backtesting engine (see more info about the engine here). This new feature provides a look at various timeframes from 1-Week to 1-Year across the selected lookback period for the backtest.

When you run a backtest now, you will see the backtest results in a tab called Backtest Statistics which contains specifics statistics for the parameters that you chose for the backtest. Another tab, called Multi-Timeframe Results,  provides a look at the backtest across multiple timeframes from 1-Week to 1-Year. Note: the multi-timeframe summary will always include overlapping results regardless of whether you select to exclude them on your backtest.

A few screenshots showing the new backtest results page are below.

backtest results

Backtest Statistics

Multi-Timeframe Results

Multi-Timeframe Results


2017-07-31 | Jason Goepfert | Comments

As mentioned last week, we’ve started a private Twitter feed for premium subscribers, and so if you’re reading this post as it’s published, you have access to that feed at @SenTrader_Prem. Just request to follow that account on Twitter and send us an email at admin “at” sentimentrader.com to notify us of your Twitter handle so we can accept it.

It’s a much better platform for much of what I personally look at during a regular trading day rather than longer-form Notes like these. Plus, I’m conscious of plugging up your email inbox.

Today’s moves, in the Dow, anyway, are dominated by Boeing and it’s worth noting some of the developments around that. First, the moves are pushing the Dow to its fourth straight 52-week high, even though other indexes like the Russell 2000 are suffering their fourth straight decline. That’s never happened before. There were three times it went 3 days in a row (NOTE: the following charts are pulled from the premium Twitter feed and are not optimized for this format so the might appear a bit blurry).

Again, the Dow’s streak is being triggered by BA, which has accounted for 57% of the Dow’s advance over the past month. That’s 322 of the Dow’s 563 point gain.

The move in BA is historic. Since 1980, it has managed to rally 15% in a week and move to a new 52-week high during only one other time period – August/September 2000.

Options traders have taken note, trading twice as many calls as puts.

Because that one stock is causing so much of the Dow’s advance, there is a structural crack in its rally. There have been 8 Hindenburg Omen signals triggered in the Dow over the past 30 days, matched only one other time since 1999. That was in October 2000.

The run in BA is historic, so who knows if it’s about to take a breather. Seems like it should be. But the broader Dow index is vulnerable because of its reliance on that single stock that is pushing its historic boundary.


2017-07-24 | Eric Brown | Comments

Over the weekend, we added a new backtesting feature to the site.

The current version of the backtesting engine is fairly straightforward. You select an symbol (index, commodity or ETF) and then select the indicator and indicator parameters that you want to backtest.   All fields are required except for the “Indicator Smoothing” dropdown and “exclude Overlapping Observations” checkbox.

A screenshot of the backtest engine setup screen is below.

sentimenTrader backtest engine

When you run the backtest, you’ll see a backtest results screen that provides a summary for the backtest parameters and results for the backtest (include the individual trade specifics).  A sample backtest result for QQQ using QQQ Social Sentiment is provided below.

QQQ Social Sentiment Backtest

 

We are planning some enhancements to the backtesting engine in the near future to make it more powerful and user friendly. A few of these planned enhancements are:

  1. Add a chart to the results page to display index, indicator and trade information
  2. Add more advanced trade statistics (annualized return, average win %, etc)
  3. Add a filter for the index to use to filter backtests (e.g., only consider a ‘trade’ if price is above 200-day moving average)
  4. Provide more insight into the symbol / indicator prior to running backtest (e.,g display chart and info about symbol / indicator when selecting parameters).

 


2017-07-13 | Jason Goepfert | Comments

This afternoon, in association with Bespoke Investment Group and a few other services, we will be rolling out a Premium Twitter feed that interested accounts can subscribe to outside of a regular SentimenTrader subscription.

If you are an existing SentimenTrader Premium tier subscriber, this is available to you at no extra charge. Because of how Twitter handles its user data, there is currently no way to automatically link Twitter to our membership management system, so:

If you wish to follow our Premium Twitter account (@SenTrader_Prem), please request to follow that account on Twitter. You’ll then need to send an email to admin@sentimentrader.com with your name and Twitter handle so that we know you’re already a Premium subscriber.

If you’re not currently a Premium tier subscriber, then you can subscribe to the feed through a company called Premo that partnered with the good folks at Bespoke Investment Group to create a way to manage private Twitter feeds. The cost for the feed itself is $10/month and is handled separately from your regular SentimenTrader subscription. Click here for more info.

Again, if you have a Premium tier subscription, you do NOT need to subscribe to the Twitter feed separately.

That Twitter feed is a more natural extension of what we’re looking at throughout the day than these Notes are. We’re conscious of clogging up your email inbox with Note after Note, so Twitter is a much better platform for quick looks throughout the day.

Also, you’ll see a number of results from the “sentimenTrader Backtest Engine”. Eric is building out a feature to allow you to backtest almost all indicators that we show on the site, and as Premium members you’ll be able to test the full history of the indicators (regular tier members will be limited to two years of history). It’s in the final testing stages and we should be able to roll it out in the next week or two. There are some useful enhancements in store in the weeks after that, as well. I’m personally excited as it’s been extremely helpful already.


2017-07-10 | Jason Goepfert | Comments

Over the past couple of sessions, there have been some notable moves in silver, which seems like it should be a good sign for the metal going forward. There was heavy selling to end the week, then a reversal from a low on Monday. That often signals exhausted selling pressure.

Unfortunately, it hasn’t been so easy for silver. Prior times when it behaved like this, it led to inconsistent results, even a bit weaker than average on some time frames.

In the coming days, we’ll be rolling out a private Twitter feed for premium users, which is an excellent platform for much of what we look at during the day, as opposed to longer-form posts that clog up your email inbox. We should be able to give full details on how to access the feed on Tuesday or Wednesday.

As part of fleshing that out, we posted a couple of studies related to the move in silver, which show just how inconsistent it’s been when exhibiting potential signs of selling exhaustion.

One good sign is that money managers have been giving up on it, so that their proportion of long to short positions is near the lowest levels in several years.

That’s one of the few bright spots, though. Even Monday’s reversal isn’t very intriguing. Other times it reversed from a low, it once again led to wildly different results. That’s not what we look for when trying to find an edge.

Silver has had a better chance at bottoming when our Optimism Index is below 30 and we’re not quite there yet (it closed at 36 on Friday). If it keeps falling in the coming weeks and starts to trigger more signs of excessive pessimism, we’d probably see more consistent signals of exhausted selling pressure than we’re seeing now.


next page →